Floating-point–consistent cross-verification methodology for reproducible and interoperable DDA solvers with fair benchmarking
Clément Argentin, Patrick C. Chaumet, Michel Gross, Maxim A. Yurkin · 2026
Read Floating-point–consistent cross-verification methodology for reproducible and interoperable DDA solvers with fair benchmarking by Clément Argentin, Patrick C. Chaumet, Michel Gross, Maxim A. Yurkin (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-02871.txt.
Description
Clément Argentin, Patrick Chaumet, Michel Gross, and Maxim Yurkin make three rival light-scattering codes agree to the last digit. Their 2026 paper benchmarks the discrete dipole approximation, a method for computing how oddly shaped objects scatter light. Independent codes differ in conventions, settings, and defaults, so fair comparisons were nearly impossible.
Their software-assisted method aligns every free parameter across DDSCAT, ADDA, and IFDDA, reaching machine-precision agreement. Practical equivalence tables let users reproduce each other's runs across codes. They then compare speed on central and graphics processors under several parallel schemes. The package also supports regression testing and exact reproducibility checks for future releases.
Catalog record
| CREATOR | Clément Argentin, Patrick C. Chaumet, Michel Gross, Maxim A. Yurkin |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 111252 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2603.02871